Skip to main content

Full text of "NASA Technical Reports Server (NTRS) 19890010158: Halogen Occultation Experiment (HALOE) optical filter characterization"

See other formats


:• -V /r * ?V * '-i ■ • ' •• V •• • ;,'C-. ' . ••..■_• •• V. ' : V ’ 

Halogen Occultation 
Experiment (HALOE) 
Optical Filter 
Characterization 


( t A S i ~ i I'i " 1 u ^ ) 1 ^ 

i. > t r. ri f E i'i (E^i-E) 
i. i f i-^C iti- ii.? i ■ 


icuki ccunnics 

u .A-U I 3 Ilfcf 

i v A i l i * fc 1 




1 4 L 


V, 0 / 3 > 


Gale A. Harvey 


APRIL 1989 



wgc-l 2 S 


U tic let ~ 
t1£5i c . b 




NASA Technical Memorandum 4104 


Halogen Occultation 
Experiment (HALOE) 
Optical Filter 
Characterization 


Gale A. Harvey 
Langley Research Center 
Hampton, Virginia 


NASA 

National Aeronautics and 
Space Administration 

Office of Management 

Scientific and Technical 
Information Division 


1989 


Contents 


Summary 

1. Introduction 

1.1. Purpose of H ALOE 

1.2. HALOE Measurement Technique 

2. Filter Descriptions 

3. Measurement Program 

4. Measurement Techniques and Instrumentation .... 

4.1. FTIR Spectrometers and Thermal Housings . . . 

4.2. Interferometer Operating Parameters 

5. Measurement Results 

5.1. In-Band Transmissions 

5.2. Spectral Shift and Throughput Versus Temperature 

5.3. Out-Of-Band Transmissions 

5.4. Reflectivity 

5.5. In-Band Time Stability 

6. Concluding Remarks 

References 

Tables 

Figures 



in 


ms. fww nsf 


Summary 

The Halogen Occupation Experiment (HALOE) 
is a solar occupation instrument that will fly on 
the Upper Atmosphere Research Satellite (UARS) 
to measure mixing ratio profiles of O 3 , HF, HC1, 
CH 4 , NO, H 2 O, and NO 2 in the stratosphere and 
lower mesosphere. The associated atmospheric pres- 
sure profile will be inferred from absorption measure- 
ments in a primary CO 2 band. The inversion of the 
HALOE data will be critically dependent on a de- 
tailed knowledge of the eight optical filters used in 
the HF, HC1, CH 4 , and NO gas correlation chan- 
nels and in the O 3 , H 2 O, N0 2 , and CO 2 radiometer 
channels. 

In order to meet the stringent HALOE require- 
ments, a filter characterization program was 
undertaken to measure in-band transmissions, out- 
of-band transmissions, in-band transmission shifts 
with temperature, reflectivities, and in-band trans- 
mission stabilities with time of three sets (flight, envi- 
ronmental witness, and spectral witness) of the eight 
filters. The measurement results of this program are 
presented herein. The in-band transmissions of the 
corresponding filters of the three sets differ signifi- 
cantly. Five of the eight filters (NO, O3, H2O, NO2, 
and CO2) have measurable out-of-band leaks in the 
spectral range from 2 to 20 ^m, but the leaks are not 
of consequence to HALOE; also, the spectral shift 
with temperature varies from 0.02 cm~ 1 /°C for the 
NO2 filter to 0.4 cm _1 /°C for the HF filter. Much 
spectral structure is present in the reflection spectra, 
and no change with time has yet been measured in 
the in-band transmission. 

The three unexpected results that are of possi- 
ble significance to HALOE are (1) in-band trans- 
missions of corresponding filters from different sets 
differ measurably; (2) out-of-band, Fabry-Perot in- 
duced transmissions are orders of magnitude larger 
than the conventional out-of-band transmissions of 
the two-element filters, and this could affect complex- 
precision optical experiments such as the Limb In- 
frared Monitor of the Stratosphere (LIMS); and 
(3) filter throughput changes by approximately 
5 percent over the expected operating temperature 
range of HALOE. 

1, Introduction 

1.1. Purpose of HALOE 

The Halogen Occultation Experiment (HALOE) 
is a solar occultation experiment that will fly on 
the Upper Atmosphere Research Satellite (UARS) 
and will measure, on a global basis, the concentra- 
tion profiles of a number of trace gas constituents 


in the stratosphere (ref. 1). The species to be mea- 
sured are O 3 , HC1, HF, NO, CH 4 , N0 2 , and H 2 0. 
(In order to correlate the data as a function of at- 
mospheric pressure, the C0 2 transmittance profile 
is also measured.) These species were selected be- 
cause they will permit the scientific community to 
study stratospheric ozone depletion resulting from 
chlorine in the stratosphere and to determine the 
relative amounts of chlorine from natural and man- 
made sources. The possibility of ozone depletion re- 
sulting from man-made chemical compounds enter- 
ing the atmosphere has been a major concern for 
the past two decades (e.g., refs. 2 to 5). The con- 
cern over the particular effects of chlorine released 
from man-made chemical compounds (especially the 
fluorocarbons CFCI 3 and CF 2 C1 2 ) used as refriger- 
ants, cleaning compounds, and foaming agents has 
surfaced more recently (refs. 6 and 7). It is this lat- 
ter concern that HALOE will address directly. The 
data derived from HALOE will, however, permit ex- 
tensive studies of stratospheric chemistry as a whole 
and, in particular, will permit studies of the inter- 
actions between the oxides of nitrogen (NO x ), chlo- 
rine (ClOx), hydrogen (HO x ), and their overall effect 
on stratospheric ozone. 

The HALOE optics are shown schematically in 
figure 1 and consist of 116 discrete elements, which 
include mirrors, lenses, beam combiners and split- 
ters, windows, spectral filters, neutral-density filters, 
and detectors. All the optical components are at- 
tached to the optical bed with separate mounts that 
permit individual alignment. The majority of the op- 
tical elements are fabricated from germanium. The 
optical bed is fabricated from aluminum. 

1.2. HALOE Measurement Technique 

The HALOE instrument measures the atmo- 
spheric absorption of solar radiation in eight chan- 
nels in the spectral range from 5000 to 1000 cm ” 1 
(2 to 10 fnn) during both sunrise and sunset oc- 
cultation events. (Values of the spectral range are 
given in two sets of units, or interchangeably, when 
considered useful.) By measuring the solar radia- 
tion with and without the intervening atmosphere 
during a solar occultation, the absorption of solar 
infrared radiation by the atmosphere can be deter- 
mined and the concentration of specified atmospheric 
trace gases can be calculated. The HALOE instru- 
ment uses both conventional optical filter radiome- 
ters (O 3 , H 2 0, NO 2 , and C0 2 ) and gas-filter cor- 
relation radiometers (HC1, HF, CH 4 , and NO). Gas 
filters are used when a high degree of specificity is 
required (i.e., when the gas is of very low concentra- 
tion and the spectral region is strongly contaminated 
with interfering gases). In the gas- filter concept, a 


specified quantity of the gas of interest, at a known 
pressure and temperature, is placed in a correlation 
gas cell in the optical path of the instrument. The 
correlation of the atmospheric absorption and the 
spectral signature of the gas in the cell in the band 
pass of the corresponding optical filter provides a sig- 
nal that can be interpreted to obtain the atmospheric 
concentration of that gas (ref. 1). 

The application of a gas-filter approach to satel- 
lite applications is relatively new. The technique 
was pioneered at the University of Oxford by the 
application of the Selective Chopper Radiometer 
(SCR) (ref. 8) and the Pressure Modulation Ra- 
diometer (PMR) (ref. 9) on the Nimbus satellites. 
The technique was also employed by the NASA Lang- 
ley Research Center (LaRC) for the Measurement 
of Air Pollution from Satellites (MAPS) experiment 
which measured global tropospheric carbon monox- 
ide during the second Space Shuttle mission (ref. 10). 
The PMR, MAPS, and HALOE approaches differ 
in that the PMR instrument modulates the pres- 
sure of the gas in the filter cell, whereas in MAPS 
and HALOE the total pressure of the gas in the 
filter cell is fixed and fluctuates only over a nar- 
row range because of natural temperature variations. 
The PMR and MAPS measurement approaches both 
differ from that of HALOE in that PMR is a limb- 
viewing instrument and MAPS is a nadir-viewing in- 
strument, whereas HALOE takes data during solar 
occultations. 

Interpretation of the gas-filter channel signals, as 
well as those of the conventional filter radiometers, 
requires accurate characterization of the optical fil- 
ters in order to retrieve good determinations of the 
atmospheric trace gases of interest. The design, fab- 
rication, and characterization of 15-pm CO 2 filters 
for the Selective Chopper Radiometer is contained in 
reference 8. 

The full range of signal change due to the gas 
of interest in the gas correlation channels will be of 
the order of 1 percent, but this 1 percent is ampli- 
fied and digitized by a 12-bit analog-to-digital con- 
verter (ADC). Thus, the signal measurement reso- 
lution over the full band pass is of the order of 10 5 . 
This high measurement resolution leads to a 
10 -3 percent transmission- measurement requirement 
over 10-em _1 intervals for out-of-band spectral re- 
gions that can be seen by HALOE detectors. In 
order to meet the HALOE requirements for precise 
information of the eight optical filters, a filter char- 
acterization program was undertaken to measure the 
in-band transmissions, out-of-band transmissions, in- 
band transmission shifts with temperature, reflectivi- 
ties, and in-band transmission stabilities of three sets 
of the eight filters. 


An optical schematic showing the location of the 
filters is presented as figure L The measurement 
apparatus and techniques are described in the text. 

2. Filter Descriptions 

Of the eight filters, three sets (flight, spectral wit- 
ness, and environmental witness) were used in the fil- 
ter characterization program. (That is, a total num- 
ber of 24 filters were used.) The flight filter set was 
mounted in flight filter holders and handled in ac- 
cordance with clean room procedures. The spectral 
witness filters were similar to the flight filters but 
were mounted in 2.0-in-diameter brass filter hold- 
ers. The environmental witness filters were similar 
to the spectral witness filters except that the ra- 
diometer filters had a 1.1-in. diameter rather than a 
0.8-in. diameter. 

The approximate center frequency, substrate ma- 
terial, and diameters of the HALOE optical filters 
are listed in table I. The germanium elements were 
the in-band transmission elements, and the MgF 2 
and ZnS elements were long-wavelength blocking el- 
ements for the last three filters listed in table I. 

All the optical filters, except the NO 2 channel fil- 
ters, were obtained from Optical Coating Laboratory, 
Inc. (OCLI). The NO 2 channel filters were obtained 
from Balzers Optical Group. As will be shown later, 
the NO 2 filters have a different band pass shape and 
spectral shift with temperature behavior than the 
OCLI filters. All the filters for a given gas channel 
are believed to have been coated at the same time. 
The detectors for the four radiometer channels are 
bolometers sensitive to radiation greater than 10 pm, 
and separate elements were needed to block the high 
transmission of germanium from 400 to 800 cm -1 . 

A photograph of a gas correlation-channel filter 
is presented as figure 2. The filters had a nominal 
wedge of 12 min of arc to minimize spectral chan- 
neling and a nominal thickness of 0.04 in. The two- 
element filters were separated by 0.15 in. 

3. Measurement Program 

The filter characterization program initially con- 
sisted of five types of measurements: in-band trans- 
mission, out-of-band transmission, spectral shift with 
temperature, reflectivity, and in-band transmission 
stability with time (the aging effect). Later in the 
program, filter throughput was also measured as a 
function of temperature. The optical filters in the 
HALOE instrument are all in parallel light beams; 
therefore, the ambient in-band transmission in a 
parallel beam was required. Accuracy and linear- 
ity were also important parameters for the in-band 
measurements. Repeatability was a primary param- 
eter in the comparisons of filters of the same set. 


2 


The spectral region of principal interest for the in- 
band measurements was between the 1-percent trans- 
mission points. Transmission measurements in the 
10" 5 range were required for the out-of-band mea- 
surements. Dynamic range, transmission precision, 
and signal- to- noise ratio were primary parameters for 
these measurements. The initial requirement for out- 
of-band measurements was for the spectral interval 
from 0.5 to 50 /im. However, since germanium is 
the substrate of most of the optics in front of the 
filter (fore-optics), the shortwave limit became the 
germanium cutoff at 6300 cm" -1 (1.5 /im), as veri- 
fied by the manufacturer’s data and interferometer 
measurements. Similarly, the longwave limits be- 
came the substrate cutoff as verified by interferome- 
ter measurements of windows of the same substrate 
materials. Filter alignment and optics positions also 
became primary parameters in these measurements. 

The philosophy for handling the filters was to 
perform as many of the measurements as possible 
on the spectral and environmental witness filters in 
order to minimize handling of the flight filters. The 
in-band transmission and temperature-effects data 
were written on magnetic tape for use in modeling 
the instrument and data retrieval purposes. 

It is generally known that the in-band transmis- 
sion of an interference filter can shift with change 
in temperature. The spectral shift was obtained 
from measurements at 60°F (the nominal operat- 
ing temperature) and also at 20° F hotter and colder 
than 60°F. The filter throughput measurements were 
made at 45°F, 80°F, and 120°F. Temperature unifor- 
mity and stability were primary parameters for spec- 
tral shift and throughput measurements. The reflec- 
tivity of the filters was obtained by replacing a gold 
reference mirror in the parallel-beam accessory with 
the filter. Thus, the reflectivity measurements are 
all referenced to a gold mirror. These measurements 
covered the spectral range from 800 to 6300 cm" 1 , 
The age stability of the filters is monitored by mea- 
suring the in-band transmission of spectral witness 
filters beginning in 1986. The measured transmis- 
sion in a long-path interferometer is extremely sensi- 
tive to filter orientation at precision levels of tenths 
percent transmission, primarily from beam steering 
of the wedged filter. The optical alignment was a 
major parameter for the reflectivity and age-stability 
measurements. 

4. Measurement Techniques and 
Instrumentation 

The basic technique used to measure the trans- 
mission of the filters was to obtain the ratio of a 
single-beam spectrum from a Fourier transform in- 
frared (FTIR) spectrometer, passed through the fil- 


ter, to a single-beam spectrum obtained under the 
same conditions except that the filter was removed 
from the optical path. Two similar, moderate- 
resolution (16 to 0.06 cm -1 ) FTIR spectrometers 
(Nicolet 7199 and Nicolet 170SX) were used for most 
of the filter measurements. A Nicolet 740 FTIR spec- 
trometer was used for some of the measurements from 
400 to 800 cm" 1 . FTIR operating parameters (res- 
olution, number of scans averaged, spectral interval, 
sample spacing, prefilters, aperture, electronic filters, 
and gain settings) were varied over wide ranges de- 
pending upon which HALOE filter and filter charac- 
teristic were being measured. 

A 2-year filter-measurement feasibility study was 
conducted during the 1983-84 time period. Tests 
were performed on different spectrometers, includ- 
ing dispersive infrared (IR) spectrometers, using dif- 
ferent measurement techniques, i.e., ratioing inter- 
ferograms for out-of-band leaks. Since optical filter 
measurements of this scope had not previously been 
performed at LaRC, accessories and support equip- 
ment were acquired, designed, fabricated, and tested. 
A class 100 clean room air filter was installed over 
the 170SX spectrometer so that the filters could be 
handled in a lint- free environment. 

4.1. FTIR Spectrometers and Thermal 
Housings 

Two FTIR spectrometers were used: one for the 
spectral interval from 1800 to 6300 cm" 1 (CaF 2 
beamsplitter and InSb detector), and one for the 
spectral interval from 800 to 2000 cm" 1 (KBr beam- 
splitter and MCTA detector). Each of the spec- 
trometers was equipped with its own microprocessor 
and peripherals, including plotter and tape recorder. 
Each spectrometer optical table was fully enclosed 
with a hood and was purged with nitrogen obtained 
from the boiloff of liquid nitrogen. The gas bearings 
of the spectrometer moving-mirror carriage were also 
supplied with boiloff liquid nitrogen. The temper- 
ature of the spectrometer optical tables was moni- 
tored with digital readout temperature and humid- 
ity gauges. The spectrometers were located in a 
laboratory with dedicated temperature and humid- 
ity control. The FTIR optical bench from 1800 to 
6300 cm" 1 rests on an air-bladder shock-absorbing 
system. 

A special housing was obtained for the temper- 
ature characterization of the filters. The housing 
consisted of a 2- by 3-in. baseplate that could fit 
into the FTIR sample stand. The baseplate was 
drilled and fitted to allow fluid to flow through 
the baseplate. A cylinder with 1.5-in. inside di- 
ameter (I.D.) and 2.0-in. outside diameter (O.D.), 
internally wound with resistance-heating wire, was 


3 


attached to the baseplate. A coverplate with a 
0.75-in-diameter opening fitted over the cylinder op- 
posite the baseplate. The housing was wrapped 
with closed-cell, polystyrene plastic foam for ad- 
ditional thermal insulation. For the colder-than- 
ambient filter measurements, the resistance-heating 
cylinder was later replaced with a brass cylinder of 
1.7-in. I.D. Later, a small heater button fastened 
to the filter holder was used for the throughput 
measurements. Calibrated thermistors were potted 
to the filter holders and were read with 5V2 digit 
multimeters. 

4.2. Interferometer Operating Parameters 

The two commercial FTIR spectrometers used in 
the filter characterization program are versatile in- 
struments. The spectral resolution is determined pri- 
marily by the distance of travel of the moving mir- 
ror. Resolutions of 0.24 cm -1 were used for most of 
the in-band transmission measurements, and 4 cm" 1 
resolution was used for most of the out-of-band mea- 
surements. The number of interferograms that were 
averaged was typically 16 for in-band transmission 
measurements and 1024 for out-of-band measure- 
ments. Optical prefilters, typically 500 cm -1 wide, 
were used for in-band measurements and to inves- 
tigate some out-of-band leaks. Source apertures of 
1.1 mm, 2.3 mm, and 6.4 mm were available, with 
2.3 mm being used for the in-band measurements 
and 1.1 mm and 6.4 mm being used for the out- 
of-band measurements. Amplifier gains of 1 were 
generally used for in-band measurements and of 1 
and 128 were used for out-of-band measurements. 
The sample spacing (nonoverlapping spectral inter- 
val) was typically eight for the in-band transmission 
and two for the out-of-band transmissions. The elec- 
tronic (noise rejection) filters were chosen to opti- 
mize the signal-to-noise ratio for each measurement, 
as was the moving-mirror velocity. A number of soft- 
ware parameters were also chosen to optimize the 
results. Tests were conducted to verify the effects 
of the measurement parameters, i.e., aperture, gain, 
prefilters, electronic filters, mirror velocity, number 
of scans averaged, and interferometer temperature 
and temperature stability. The parameters that gave 
the greatest signal-to-noise ratio and the best linear- 
ity for each particular measurement were generally 
used. 

5. Measurement Results 

The optical filter measurement program provided 
detailed and specific information on the spectral 
transmissions of the HALOE optical filters. Certain 
optical characteristics such as angular deviation, ho- 
mogeneity, and angle tuning (spectral tuning) were 


not measured. The in-band transmissions and the 
out-of-band filter leaks received the most attention 
and are believed to be state-of-the-art infrared filter 
measurements. Very precise measurements were also 
made of the change in filter throughput at several 
different filter temperatures. 

5.1. In-Band Transmissions 

The in-band transmissions of the HALOE optical 
filters are presented as figures 3 to 10. Each filter 
has been measured several times at ambient tem- 
peratures (approximately 25°C), and the precision 
of the measurements is indicated by the nearly over- 
lapping curves for a given filter. However, very mea- 
surable differences in the transmission spectra of the 
individual filters of the same gas channel are clearly 
seen. These measurements were made with an //4 IR 
beam. A six-mirror accessory was placed in the sam- 
ple compartment of the interferometer in order to 
collimate the // 4 IR beam for flight-filter in-band 
measurements. Since the filters were wedged to pre- 
vent channeling, the absolute transmission, as mea- 
sured by the interferometer, is a function of filter ori- 
entation when placed in a collimated beam because 
of the beam steering introduced by the wedge. Al- 
though the data were not shown here, the collimated 
beam transmissions are similar to the //4 IR beam 
transmissions. 

Precautions were taken with the collimated in- 
band transmission measurement to ensure that the 
IR beam was equally bounded for both sample and 
background spectra. This was particularly necessary 
for the O3, H2O, NO2, and CO2 filters with clear 
apertures of 0.7 in. The full aperture could not be 
filled because of vignetting by the accessory mirrors 
and mounts and the prefilter mount. A 0.4-in. beam 
aperture was used, and an empty filter holder was 
inserted in the filter position for the background 
spectrum for this purpose. Alignment was facilitated 
with the interferometer HeNe laser beam. 

All the filters, except the NO2 channel filters, 
were made by Optical Coating Laboratory, Inc. 
(OCLI). Figures 3, 4, and 6 to 10 show that the 
OCLI filters all tend to have flat tops and very steep 
sides. The NO2 channel filters were made by Balzers 
Optical Group and have a Gaussian shape. 

5.2. Spectral Shift and Throughput Versus 
Temperature 

The HALOE filter characterization program ini- 
tially required spectral shift measurements of the four 
flight radiometer filters and four spectral witness, 
correlation-channel filters at 40°F, 65°F, and 80°F 
(the nominal operating range of the instrument). 
Later, when significant differences were observed in 


4 


in-band transmission between spectral witness and 
flight filters, temperature measurements were also 
made on the flight correlation-channel filters. 

A spectral shift of the in-band transmission of 
the flight filters can be seen in figures 11 to 18. 
The average (An/A T) shifts of the filters for the 
eight HALOE channels are presented in table II, 
where v represents the spectral frequency and T 
represents the temperature. These spectral shifts 
were calculated by measuring the spectral frequency 
at 10-percent transmission intervals (normalized to 
peak transmission equal to 100-percent transmis- 
sion). Thus, An/ A T represents averages of 19 points 
over the bandwidth of the filter. It can be seen from 
table II that for six of the filters, the spectral shift 
is generally a function of the filter frequency and in- 
creases as the frequency increases. The CO2 filter 
is an exception. However, for the NO2 filters, which 
were supplied by a different manufacturer and have 
a different in-band shape than the other seven filter 
sets, the spectral shift is quite small. 

The throughput of a filter can be defined as the 
integral of transmission over the filter bandwidth. 
During the spectral shift measurements and HALOE 
instrument testing, it was noted that the peak trans- 
mission and the bandwidth vary with temperature. 
The filters were heated to 50° C to increase the preci- 
sion of the throughput measurement. The seven fil- 
ter sets with measurable spectral shift have decreas- 
ing throughput with increasing temperature because 
both the peak transmission and the bandwidth de- 
crease with increasing temperature. A typical case is 
a decrease of 5.9 percent in throughput when the CO2 
filter temperature is increased from 25°C to 54°C. 
The throughput of the NO filter decreased by 5.6 per- 
cent when the filter temperature was increased from 
6°C to 51°C. Three transmission curves at each tem- 
perature are presented for each of the eight filters in 
figures 11 to 18. 

5.3. Out-Of-Band Transmissions 

The spectral witness filters were measured ini- 
tially for possible out-of-band transmissions (leaks). 
Later, the flight filters were also measured. The 
spectral interval covered in the out-of-band measure- 
ments was determined by the germanium cutoff for 
higher wave numbers at 6300 cm' 1 and by the fil- 
ter substrate transmission for the lower wave num- 
ber limit. For example, the CH4, HC1, CO2, and 
HF filters were all on Si02 substrate, and the out-of- 
band measurements covered the interval from 1800 
to 6300 cm” 1 . Out-of-band measurements for the 
ZnS and Ge substrates of the O3 and for the Ge sub- 
strates of the NO filter covered the interval from 400 
to 6300 cm” 1 . Out-of-band measurements for the 


MgF and Ge substrates of the H2O and NO2 filters 
covered the spectral interval from 800 to 6300 cm” 1 . 

It was found that by orienting the two-element 
filters perpendicular to the IR beam, a Fabry-Perot 
leak could be induced in these filters if a flat op- 
tic window, i.e., a prefilter, was placed in front of 
the filter. Orientation of the filter was determined 
by reflecting the HeNe laser beam of the interfer- 
ometer back upon itself. Figure 19 shows the leaks 
in the O3 filter perpendicular to the IR beam, both 
with and without a flat optic window (NaCl) for the 
mid-IR region. These leaks increase in transmission 
and broaden when placed in a Fabry-Perot cavity. 
The strong Fabry-Perot enhancement of very small 
leaks in the spectral filters is a probable cause of 
the “side lobes” and the post-mission conclusion of a 
1.5-percent leak in the CO2N filter of the Limb 
Infrared Monitor of the Stratosphere (LIMS) (ref. 11). 

Figures 20 to 27 show out-of-band transmissions 
for representative regions of the other flight filters. 
All these measurements were made with the // 4 IR 
beam. The leak in the NO filter (fig. 23) at 600 cm -1 
is of no consequence to HALOE, because two AI2O3 
windows (1500-cm -1 cutoff) in the NO channel block 
this leak. Except for the O3 and H2O filters, all 
the out-of-band data presented (figs. 22 to 27) are 
for “normal” leaks. The H2O filter was the most 
extensively studied (approximately 100 spectra). 

The “normal” out-of-band transmission measure- 
ments can be summarized as follows: 

1. The NO filter has a large leak (> 10-percent 
transmission) below 1000 cm” 1 . 

2. The CO2 filter has small leaks (<0. 1-percent 
transmission) in the interval from 4000 to 6000 cm” 1 . 

3. The NO2 filter has small leaks (approximately 
0.001-percent transmission) in the interval from 2000 
to 2500 cm” 1 . 

4. The O3 filter has small leaks (<0. 001-percent 
transmission) in the intervals from 2400 to 3900 cm” 1 
and 4400 to 6000 cm” 1 . 

5.4. Reflectivity 

The reflection measurements were made by plac- 
ing the filters in the position of one of the plane mir- 
rors in the six-mirror parallel-beam accessory. The 
interferograms are only about 25 percent as big when 
using this accessory as compared with the //4 IR 
beam because of reflection losses and image degra- 
dation from the accessory. In addition, the inter- 
ferograms and, hence, the measured reflectivity are 
extremely sensitive to the tilt of the plane mirrors 
(i.e., gold reference mirror and filter) since the an- 
gles of reflection are double the mirror tilt. The 
result is a lateral displacement of the interferom- 
eter IR source image on the detector between the 


5 


reference spectrum and the filter-reflection spectrum 
when the mirrors are in a parallel beam. These 
aspects of the reflection measurements resulted in the 
accuracy, precision, and repeatability of the reflection 
measurements being much lower than for the in-band 
and temperature-effects measurements. Fortunately, 
the reflectivity of the optical filters is not used in 
the data retrieval of HALOE and, consequently, is 
of low priority. Nevertheless, all the spectral wit- 
ness filters were measured over the spectral interval 
from 800 to 6300 cm” 1 . A typical reflection spectrum 
(of the H2O filter) is presented as figure 28. As ex- 
pected, the measured reflectivity of the filters is low 
(approximately 10-percent reflectivity) in the band 
pass regions. However, the filters also have numer- 
ous other spectral regions of low reflectivity. That 
is, there is a lot of spectral structure in the reflection 
spectrum. This structure is somewhat similar to the 
structure seen in the Fabry- Perot leaks and probably 
contributes to these leaks. 

5.5. In-Band Time Stability 

The initial filter stability-with-time measure- 
ments were performed on the spectral witness filters. 
The filter stability was by comparative measurements 
of the peak transmission of the filters in the // 4 IR 
beam. The measurements were initially scheduled 
to be made every 6 months, but conflicting data re- 
quirements of the interferometers precluded adher- 
ence to this schedule. Since the filters are wedged to 
minimize spectral channeling, the IR beam is devi- 
ated when passing through the filter. Even though 
the beam in the sample compartments is //4 and is 
reduced to a diameter of about 6 mm at the cen- 
ter of the compartment, there is no image position 
where the filter-induced deviation of the IR beam is 
completely compensated for by the downstream op- 
tics with no translation of the source image on the 
detector. Consequently, the rotational alignment of 
the filter has an effect on the measured transmission 
at a 1-percent transmission level. The understand- 
ing of measurement repeatability in transmission 
was aided by a concurrent program of monitoring 
condensable-volatile-material contamination of op- 
tics at a 0. 01-percent transmission level in the 
HALOE clean room using the Nicolet 170SX spec- 
trometer (ref. 12). The transmission repeatability 
is also affected by interferometer alignment and fil- 
ter location. Transmission of a CH4 filter at two 
different times is presented as figure 29. 

Although good baseline transmissions were not 
obtained early in the filter program, stability mea- 
surements with improved filter orientation and posi- 
tion indexing will be performed on a set of 


“instrument- build” filters on a periodic basis. There 
is no indication of change in transmission over a 
2-year period within the estimated precision (5 per- 
cent) of the measurements. However, a similar CO 
spectral filter (fig. 30) from the MAPS program 
(ref. 10) shows extensive and continuing peeling and 
flaking even though maintained in an environmen- 
tally controlled and protected area. The half-width 
of the flaked area of the filter is about four times the 
half-width of the unflaked area. 

6. Concluding Remarks 

The Halogen Occultation Experiment (HALOE) 
is a solar occultation instrument that will fly on the 
Upper Atmosphere Research Satellite (UARS). The 
inversion of the HALOE data will be dependent on 
a detailed knowledge of the optical filters used in 
the HF, HC1, CH4, and NO gas correlation chan- 
nels and in the CO2, NO2, H2O, and O3 radiometer 
channels. A 2-year, measurement-feasibility study 
was conducted to define and determine the best way 
to perform the required filter characterization. The 
filter characterization was performed in-house using 
two moderate-resolution Fourier transform infrared 
spectrometers to perform various measurements over 
the spectral interval from 400 to 6300 cm” 1 (25 
to 1.6 fim). The primary measurements were in- 
band transmission, throughput change with temper- 
ature, and out-of-band leaks. Measurements were 
also made of in-band stability with time and reflec- 
tivity. The in-band measurements are believed to 
have greater precision than any previously obtained 
and graphically show individual differences in filters 
from the same filter set and manufacturing run. The 
out-of-band measurements were generally made at 
the 10” 4 percent transmission level and are believed 
to be of higher resolution than any similar out-of- 
band transmission measurements. These measure- 
ments have revealed a number of narrow leaks at 
the 10“ 2 percent transmission level and some greatly 
enhanced (several orders of magnitude) leaks of the 
two-element filters when placed in a Fabry-Perot cav- 
ity. The temporal stability and reflection measure- 
ments are strongly influenced by optical steering in 
the interferometer. 

The optical filters procured and selected for the 
HALOE flight instrument are well-suited, in terms of 
in-band transmission and normal out-of-band rejec- 
tion, for meeting the HALOE scientific objectives. 

NASA Langley Research Center 
Hampton, VA 23665-5225 
February 17, 1989 


References 

1. Russell, James M., Ill; Park, Jae H.; and Drayson, 

S. Roland: Global Monitoring of Stratospheric Halo- 

gen Compounds From a Satellite Using Gas Filter Spec- 
troscopy in the Solar Occult at ion Mode. Appl. Opt, 
vol. 16, no. 3, Mar. 1977, pp. 607-612. 

2. Man’s Impact on the Global Environment Report of the 
Study of Critical Environmental Problems (SCEP). MIT 
Press, c.1970. 

3. Inadvertent Climate Modification: Report of the Study of 
Man’s Impact on Climate (SMIC). MIT Press, c.1971. 

4. Grobecker, A. J.; Coroniti, S. C.; and Cannon, R. H., Jr.: 
Report of Findings — The Effects of Stratospheric Pollu- 
tion by Aircraft. TST-75-50, U.S. Dep. of Transportation, 
Dec. 1974. (Available from DTIC as AD A005 458.) 

5. Halocarbons: Effects on Stratospheric Ozone. National 
Academy of Sciences, 1976. 

6. Stolarski, R. S.; and Cicerone, R. J.: Stratospheric Chlo- 
rine: A Possible Sink for Ozone. Canadian J. Chem., 
vol. 52, no. 8 (pt. 2), Apr. 15, 1974, pp. 1610-1615. 

7. Molina, Mario J.; and Rowland, F. S.: Stratospheric 
Sink for Chlorofluoromethanes: Chlorine Atom Catalysed 
Destruction of Ozone. Nature , vol. 249, no. 5460, June 28, 
1974. 


8. Abel, P. G.; Ellis, P. J.; Houghton, J. T.; Peckham, 
G.; Rodgers, C. D.; Smith, S. D.; and Williamson, 
E. J.: Remote Sounding of Atmospheric Temperature 
From Satellites. II. The Selective Chopper Radiometer 
for Nimbus D. Proc. Royal Soc. London, ser. A, vol. 320, 
no. 1540, Nov. 24, 1970. pp. 35-55. 

9. Taylor, F. W.; Houghton, J. T.; Peskett, G. D.; Rodgers, 
C. D.; and Williamson, E. J.: Radiometer for Remote 
Sounding of the Upper Atmosphere. Appl. Opt., vol. 11, 
no. 1, Jan. 1972, pp. 135-141. 

10. Reichle, Henry G., Jr.; Connors, Vickie S.; Holland, J. 

Alvin; Hypes, Warren D.; Wallio, H. Andrew; Casas, 
Joseph C.; Gormsen, Barbara B.; Saylor, Mary S.; 
and Hesketh, Wilfred D.: Middle and Upper Tropo- 

spheric Carbon Monoxide Mixing Ratios as Measured by 
a Satellite- Borne Remote Sensor During November 1981. 
J. Geophys. Res., vol. 91, no. D10, Sept. 20, 1986, 
pp. 10,865-10,887. 

11. Gille, John C.; and Russell, James M., Ill: The Limb In- 
frared Monitor of the Stratosphere: Experiment Descrip- 
tion, Performance and Results. J. Geophys. Res., vol. 89, 
no. D4, June 30, 1984, pp. 5125-5140. 

12. Harvey, Gale A.; and Raper, James L.: Halogen Occul- 
tation Experiment (HALOE) Optical Witness-Plate Pro- 
gram. NASA TM-4081, 1989. 


7 









Transmittance, percent 


L-87-6231 


Figure 2. Photograph of a gas correlation-channel filter. 



Figure 3. Transmissions of two O3 filters. 


10 


Transmittance, percent Transmittance, percent 



1500 1504 1508 1512 1516 1520 1524 1528 1532 1536 

Wave number 


Figure 4. Transmissions of three H 2 O filters. 



1570 1575 1580 1585 1590 1595 1600 1605 1610 1615 

Wave number 


Figure 5. Transmissions of three NO 2 filters. 


11 














Transmittance, percent Transmittance, percent 












Transmittance, percent Transmittance, percent 



Figure 15. Throughput of CH 4 flight filter. 



Figure 16. Throughput of HC1 flight filter. 


17 
















smittance, percent 






percent . 








Tra 



Wave number 

Figure 29. CH 4 filter- measured transmission at two epochs. 



L-88-9115 

Figure 30. Photograph of deteriorated MAPS filter. 


24 


Report Documentation Page 

Space Administration 


1. Report No. 2. Government Accession No. 

NASA TM-4104 


4. Title and Subtitle 

Halogen Occuitation Experiment (HALOE) Optical Filter 
Characterization 


7. Author(s) 

Gale A. Harvey 

9. Performing Organization Name and Address 

NASA Langley Research Center 
Hampton, VA 23665-5225 


3. Recipient’s Catalog No. 


5. Report. Date 

April 1989 


6. Performing Organization Code 


8. Performing Organization Report No. 

L-16538 


10. Work Unit No. 

678-12-04-07 


11. Contract or Grant No. 


12. Sponsoring Agency Name and Address 

National Aeronautics and Space Administration 
Washington, DC 20546-0001 


13. Type of Report and Period Covered 
Technical Memorandum 


14. Sponsoring Agency Code 


! 15. Supplementary Notes 


16. Abstract 


The Halogen Occuitation Experiment (HALOE) is a solar occuitation instrument that will fly 
on the Upper Atmosphere Research Satellite to measure mixing ratio profiles of O3, H2O, NO2, 
NO, CH4, HC1, and HF. The inversion of the HALOE data will be critically dependent on a 
detailed knowledge of eight optical filters. A filter characterization program was undertaken 
to measure in-band transmissions, out-of-band transmissions, in-band transmission shifts with 
temperature, reflectivities, and age stability. Fourier transform infrared spectrometers were used 
to perform measurements over the spectral interval from 400 to 6300 cm -1 (25 to 1.6 /im). 
Very high precision (10 -1 percent transmission) in-band measurements and very high resolution 
(10“ 4 percent transmission) out-of-band measurements were made. The measurements revealed 
several conventional leaks at 10 -2 percent transmission and greatly enhanced (10**) leaks of the 
two-element filters when placed in a Fabry- Perot cavity. Filter throughput changes by 5 percent 
for a 25° C change in filter temperature. 


17. Key Words (Suggested by Authors(s)) 
Infrared filter characterization 
Optical filters 
Fabry-Perot filter leaks 
Filter throughput measurements 


18. Distribution Statement 

U nclassified — U nlimi ted 


Subject Categories 35 and 18 


19. Security Classif. (of this report) 20. Security Classif. (of this page) 21. No. of Pages 22. Price 

Unclassified Unclassified 27 A03 


NASA FORM 1626 OCT 86 N A 

For sale by the National Technical Information Service, Springfield, Virginia 22161-2171 


NASA-Langley, 1989